Geekbot

Geekbot is a team collaboration tool that runs asynchronous standups, polls, and surveys directly in Slack and Microsoft Teams. The bot reaches out to team members on a schedule, collects their answers, and shares the updates and summaries with the team. Geekbot also offers a public REST API, an open-source CLI, and an MCP server so AI agents can work with the same data.

What It Is

Geekbot replaces status meetings with chat-based check-ins. Teams pick a template (such as a daily standup, retrospective, sales report, or quick check-in), customize the questions and schedule, and publish. Geekbot then collects responses and shares results automatically. Scheduled summaries highlight participation, blockers, and sentiment.

Polls and surveys cover quick team feedback, such as Yes/No decisions, employee NPS surveys, and weekly team feedback. The site lists AI-powered suggestions that auto-generate answer options, and automatic grouping of open answers into themes. Results are visualized in charts and graphs.

Connecting AI agents

The Geekbot CLI (TypeScript and Bun, MIT licensed) wraps the Geekbot API so standups, reports, polls, profile, and team data can be managed from a terminal or an AI coding agent. Every command returns a structured JSON envelope with distinct exit codes. Credentials are stored in the OS keychain, and login uses browser-based OAuth with PKCE, or an API key as an alternative for CI. The README describes plugins or skills for Claude Code, Claude Desktop, Codex CLI, Cursor, and Gemini CLI, plus other Agent Skills tools such as Windsurf and Copilot. Geekbot also documents an MCP server for querying standups, fetching reports, and posting updates in natural language.

API

The REST API at api.geekbot.com uses an API key in the Authorization header and returns JSON. Current endpoints under /v2/ cover standups, reports, polls, and out-of-office periods, with cursor pagination, opt-in field expansion, structured errors, and idempotency keys on writes. Legacy /v1/ endpoints remain supported. Requests are limited to 60 per minute per user.